© Senckenberg Museum of Natural History Görlitz · 2016ISSN 1618-1735

pp. 35–42

Hennig, Ax, and Present-Day Mainstream Cladistics, on polarising characters

Michael Schmitt

Ernst-Moritz-Arndt-Universität, Allgemeine and Systematische Zoologie, Soldmannstr. 14, 17489 Greifswald, Germany

E-mail: michael.schmitt@uni-greifswald.de

Received 19 January 2016 | Accepted 2 February 2016

Published online at www.senckenberg.de/peckiana 12 Februar 2016 | Printed version 19 February 2016

Abstract

The way in which Willi Hennig and Peter Ax substantiated hypotheses on character polarity are described and compared to the presently most common practice of polarising characters by rooting an unrooted network. Although Hennig applied the outgroup comparison method implicitly, and Ax regarded this method as the only general and reliable tool for determining character polarity, there is a crucial difference to the application of this method in present-day mainstream cladistics: Traditionally (Hennig, Ax), character polarity was assessed on a character-by-character basis, while in ‘present-day mainstream cladistics’ rooting replaced the a-priori-discussion of possible character transformations.

Keywords History of phylogenetics | outgroup comparison | character polarity | methodology | Willi Hennig | Peter Ax

11 · Februar 2016

Introduction

Since the foundation of Phylogenetic Systematics by Willi Hennig (Hennig 1950, 1966) it is agreed that ‘monophyletic’ is defined as ‘comprising a stem species and all its descendants’ and that establishing hypotheses on monophyly must be based on hypotheses on uniquely derived characters – syn- or autapomorphies. Consequently, the assessment of character polarity (‘Lesrichtung der Merkmalsreihen’, deciding on plesio- or apomorph states) is the core task in phylogenetic practice. It could, therefore, be of interest to compare how Willi Hennig, the founder, Peter Ax, the most prominent German propagator, and the practitioners of present-day cladistics, met or meet this task.

Willy Hennig (20.04.1913–05.11.1976) (Fig. 1)

As described earlier (Schmitt 2001, 2013: 131ff.), Hennig had developed the basic idea that only relatively

derived characters (or character states) can substantiate a hypothesis of closer relatedness (1936a: 552, 1936b: 170). However, how we can empirically identify such a derived state remained vague. In the Grundzüge (Hennig 1950: 172 – 178) he gave four rules for the evaluation of single morphological characters, repeated in Phylogenetic Systematics (Hennig 1966: 95ff., 1982: 98ff., see also Richter & Meier 1994): paleontological character precedence, chorological progression, ontological character precedence, and the correlation of transformation series. Interestingly, he presented these empirical tools originally not as ‘criteria’ to assess polarity, but rather neutrally as indications of closer relatedness.

The so-called ‘criterion of paleontological character precedence’ (‘so-called’ because it is strictly speaking not a logical criterion but an empirical indication) would, of course, yield information for determining character polarity, or phylogenetic (= genealogic) relationships in general, with absolute certainty if the fossil record would be complete. As this is not the case, fossils can lead to erroneous decisions, as Willmann (1989: 283) has pointed out. Nevertheless, there is a certain chance to reach the correct assessment by comparing fossil

Michael Schmitt36

PECKIANA 11 · 2016

and extant specimens. Normally, or at least mostly, the geologically older character state of a transformation series is plesiomorph as compared to younger states.

The ‘chorological progression’ means that an apomorph character state is expected to occur more frequently in populations near the edges of a distributional area than in the centre, whereas the respective plesiomorph state is said to occur close to the centre of origin of a species group with higher probability than towards the borders of the area covered by this group of species. Whether or not the underlying idea on dispersal of species and character transformation appear convincing, the whole ‘criterion’ suffers from the general drawback that the ‘centre of origin’ must be identified without reference to the putative state of the species considered, i.e. whether their character states are plesio- or apomorph. It is unclear how Hennig (or anybody else) would determine the centre of origin without circular reasoning.

‘Ontological character precedence’ starts from the fact that organisms cannot ‘close due to reconstruction’ (wegen Umbau schließen, Osche 1966: 830). Therefore,

evolutionarily more recent features can be integrated into the ontogeny of an organism by ‘terminal addition’ (Sewertsoff 1931: 266ff.). Consequently, this ‘criterion’ can be phrased ‘If of two states of a character one occurs as a transitional state - an “interphaen” (Riedl 1975: 258) - in the ontogeny in one lineage and as a final state – a metaphaen - in another lineage, while the alternative character state occurs as a metaphaen only, then the first can be regarded apomorph’. Whether or not this line of argument is accepted or not, the fundamental drawback here is the general paucity of information on the ontogenetic development of the organisms under study. The ontogenetic development of neither fossils nor collection specimens can be observed. Even the study of the ontogeny of live organisms is a limited source of relevant information, although there are examples to the contrary, as was exemplified by Rudolf Meier (1997) in a critical study using sepsid larvae (Diptera). Possibly, numerous studies on the ontogeny of organisms could yield interesting data but have not yet been exploited by phylogeneticists.

Figure 1. Willi Hennig in 1975. Painting by Adele Hornig (Oppach, Germany) after a photograph, with kind permission.

PECKIANA 11 · 2016

37Hennig, Ax, and Present-Day Mainstream Cladistics, on polarising characters

Theoretically, the ‘correlation of transformation series‘ cannot establish a ‘criterion’ on its own. Already Lundberg (1972) has pointed out that this ‘criterion’ ‘cannot be regarded as a very strong criterion nor a very useful one’ (p. 400) when he discussed the three ‘general criteria’ presented by Kluge & Farris (1969). At most, it could enhance the reliability of an assessment of character polarity. The principal weakness of this tool is that all characters in a phylogenetic analysis must be regarded as varying independently, to avoid circularity. If two or more characters co-vary necessarily, they could actually be just one single character, possibly brought about by pleiotropy.

Hennig did not only not elaborate a concise and convincing method for polarising characters, he did, moreover, insist that it was not possible to provide such a method (Hennig 1984: 47). In his opinion, taxonomic experience and familiarity with the objects of study would teach the investigator how to decide. In his practical papers, he decided mostly on the basis of plausibility (‘... es scheint mir .... durchaus plausibel’, Hennig 1972:

25), or according to ideas on evolutionary processes, e.g. switch of hosts from reptiles and birds to mammals in phlebotomine flies (1972: 24). In some cases he simply relied on the statements of other authors: ‘According to NN ....’ (Hennig 1983, frequently).

In a posthumous paper (Hennig & Schlee 1978) we read that distinguishing plesiomorph from apomorph states of a character is easy and can be seen in most cases from the wide distribution of the characters outside the investigated group of species (‘Die ... Entscheidung ist oft leicht und ergibt sich meist aus der weiten Verbreitung der Merkmale außerhalb der untersuchten Artengruppe’, p. 5). Here, Hennig (and his last assistant Dieter Schlee) implicitly apply the method of outgroup comparison. We find examples of this approach very often in Hennig’s publications. His usual argument for an assessment of character polarity is ‘in comparison with the putative sister group’ (e.g. 1968: 3), or ‘since all other subfamilies of the Psychodidae have 2 spermathecae, .... the reduction of one of them is with certainty an apomorph character in the Bruchomyiinae’, 1972: 19).

Peter Ax (29.03.1927–02.05.2013) (Fig. 2)

In his early phylogenetic papers (e.g. Ax 1965), Ax followed the conventional line of reasoning, based on the methodological framework provided by Adolf Remane (10.08.1898 – 22.12.1976, e.g. 1952). Only later, e.g. in his analysis of systematics and phylogeny of the platyhelminth subfamily Trigonostominae (1971), he adopted Hennig’s method of phylogenetic analysis. In his textbooks (Ax 1984, 1988), he described three arguments for polarising of characters (‘directional arguments for evolutionary change’, according to De Jong 1980): (1) Out-group comparison, (2) comparison with the stem-lineage of a taxon, and (3) in-group comparison. Ax stated that the outgroup comparison method, as formalised by Watrous & Wheeler (1981) is the only general and reliable method to determine character polarity. His argument was (1984: 125): If a character occurs within a putatively monophyletic group of species in alternatives, then the state also occurring outside this group is likely the plesiomorphy (‘Tritt ein Merkmal in einer mutmaßlich monophyletischen Artengruppe in Alternativen auf, so ist jener Zustand, der auch außerhalb der Gruppe vorkommt, wahrscheinlich die Plesiomorphie’). He stated that the out-group can be any taxon showing the character in question in one of the alternatives occurring in the in-group – it is not required that it is the sister taxon nor that it is shown to be monophyletic. The ‘comparison with the

Figure 2. Peter Ax, during the discussion at the 31st Phylogenetisches Symposium, Freiburg im Breisgau (Germany), 26.11.1988. Photograph taken by Hannes Paulus, with kind permission.

Michael Schmitt38

PECKIANA 11 · 2016

stem-lineage’ (p. 129) corresponds roughly to Hennig’s ‘criterion of paleontological character precedence’: If a character occurs within a putatively monophyletic group of species in alternatives, then the state occurring in the stem-lineage of that taxon is likely the plesiomorphy (‘Tritt ein Merkmal in einer mutmaßlich monophyletischen Artengruppe in Alternativen auf, so ist jener Zustand, der in der Stammlinie des Taxons vorkommt, wahrscheinlich die Plesiomorphie’). He discusses the weaknesses of this argument himself, i.e. due to the incomplete fossil record we can never be certain that a geologically older character state is plesiomorph as compared to a geologically younger, provided that they belong to the same transformation series. The latter point is especially relevant because we can only determine the putative members of the stem-lineage after conducting the phylogenetic analysis. Ax discussed as a third possible argument on p. 131 the ‘in-group comparison’ (common equals primitive) but stated on p. 132 that this argument is obsolete.

In the first volume of his opus magnum ‘Das System der Metazoa’ (1995: 27f.), he mentioned the outgroup comparison method as the only tool for polarising characters. He regarded a ‘functional adaptive analysis’ as a method that could yield possible additional considerations for the interpretation of character transformation but no proper argument for character

polarity assessment. I find it remarkable that Ax had a fully operational concept of ‘out-group’, in contrast to all other contemporary phylogeneticists in Germany (e.g. Sudhaus & Rehfeld 1992: 105). An outgroup is a group of organisms exclusively chosen for the purpose of polarising a series of character states. Nothing is said and nothing has to be stated about phylogenetic relationships.

As far as I see, Peter Ax has never produced a character matrix, at least he never published one. He performed the outgroup comparison for each character individually, and he did rarey point to a concrete outgroup taxon when listing putative evolutionary novelties (aut- and synapomorphies). Very rarely he discussed alternative relationships in his system of the Metazoa (Ax 1995, 1999, 2001). This means that he composed the trees he presented on the basis of plausible ideas on character transformation and on comparison among a limited set of taxa. Considering that the number of possible dichotomous rooted trees increases exponentially with the number of taxa in the analysis – there are three possible trees for three taxa, 15 for four, 105 for five, and 34.459.425 for ten taxa – it is clear that neither Peter Ax nor anybody else could evaluate even a small proportion of the total number of possible trees. This means that Ax based his decisions on numerous tacit assumptions, on his personal experience and preferences.

Figure 3. Covers of the three textbooks that I used as methodological sources for chapter on ‘Present-Day Mainstream Cladistics’.

PECKIANA 11 · 2016

39Hennig, Ax, and Present-Day Mainstream Cladistics, on polarising characters

Present-Day Mainstream Cladistics

Especially since the ‘transformation of cladistics’ (Nelson 1979, Nelson & Platnick 1978, 1984, Patterson 1980, Platnick 1979), the landscape of approaches of phylogenetic analysis became diverse and nearly unintelligible, in spite of the analyses of Carpenter (1987) and Ebach et al. (2008). Here, I focus on ‘mainstream cladistics’ as practised in the papers published in Systematic Zoology/Biology and Cladistics during the past 40 years. As methodological sources I use the textbooks of Kitching et al. (1998), Schuh & Brower (2009), and Wiley & Lieberman (2011) (Fig. 3) that explain how to discuss character states individually. In practice, however, ‘present-day mainstream cladists’ polarise characters exclusively by ‘outgroup addition’ (Wägele 2000, p. 169), i.e. by rooting an unrooted network. Meier (1992, 1995) discussed extensively the implications of choice of the outgroup, rooting, and refraining from a-priori-determination of character polarity.

Throckmorton (1968) developed the method of outgroup comparison that was later explicitly described by Kluge & Farris (1969). Wiley published the ‘out-group rule’ (1981: 139): ‘Given two characters that are homologous and found within a single monophyletic group, the character that is also found in the sister group is the plesiomorphic character whereas the character found only within the monophyletic group is the apomorphic character’. Strictly seen, this rule requires the assessment of monophyly of the ingroup and establishing its sister group prior to polarising the characters in question. Consequently, in the second edition (Wiley & Lieberman 2011), the ‘outgroup rule’ is slightly reworded: ‘Given two (or more) homologous character states within a group studied, the state found outside this group in close relatives is the plesiomorphic state and the character found only within the group is the apomorphic state (p. 157). Watrous & Wheeler (1981) formalised the procedure of outgroup comparison and defined the ‘outgroup criterion for polarity determination’ as ‘For a given character with two or more states within a group, the state occurring in related groups is assumed to be the plesiomorphic state (p. 5). Kitching et al. (1998) refer to Watrous & Wheeler (1981), stating that only from 1981 on a consistent set of operational rules for outgroup comparison was available.

The general rationale behind the outgroup comparison method (i.e. whether by rooting or on a character-by-character basis) is the principle of parsimony. William of Occam stated in his Summa logicae that ‘pluralitas non est ponenda sine necessitate’ (a plurality must not be supposed without necessity. Beckmann 1995: 43). William of Occam stated this principle (‘Occam’s razor’) as a general rule of thinking, starting from the assumption

that only individual things are real, and that we must aim at making the number of notions higher than the number of things only where necessary. In cladistics, the principle means that we should prefer the ‘shortest’ tree revealed by the cladistic analysis. The length of a tree is measured as the sum of transformational steps from one state of a character to another, calculated either (rarely) by a human scientist or (normally) by a computer algorithm.

In all four sources cited above, the method of outgroup comparison is explained in a way that polarity determination is done for each character separately and prior to the phylogenetic analysis. However, as Wiley and Lieberman bluntly state ‘phylogeneticists rarely polarize characters a priori these days’. Schuh and Brower (2009) report that ‘polarity was usually determined on a character-by-character basis during the data-gathering phase’ by phylogeneticists following the ‘traditional Hennigian approach’. The ‘controversial interpretation of individual character polarities led to extended discussion in papers from the 1970s and early 1980s. Computer algorithms, such as the Wagner algorithm originally described by Farris, minimize the number of character-state changes among taxa without regard to the polarity … The orientation of the network created by the algorithm is then determined by specifying one taxon to identify the root. Under this formalization of cladistics, all of the literature describing methods to determine individual character polarity really addresses a non-problem. It is only the choice of the outgroup and the position of the root that matter’ (p. 98f.).

Even if hardly any present-day cladist refers in publications to an explicit description of the outgroup comparison method, they all – with extremely few exceptions – polarise characters by rooting a posteriori, i.e. after running the respective cladistic computer programme. It is trivial to explain that a priori polarising is practically impossible when dealing with molecular sequence data exclusively. Also, all other arguments – ontogenetic, paleontological, functional-adaptive – are not applicable in these cases. But there are ‘these days’ only extremely few cladistics analyses based on non-molecular characters in which polarity is determined ‘on a character-by-character basis’.

Discussion

From all discussions of the methodology of phylogenetic analysis it becomes clear that the outgroup comparison method is the most widely accepted and most frequently applied tool for polarising characters. Even Willi Hennig often used this method, however without explicit reference

Michael Schmitt40

PECKIANA 11 · 2016

or description. Nevertheless, Hennig stressed other, possibly additional, lines of evidence for the assessment of character polarity: ontogeny, palaeontology, chorology. Peter Ax discussed extensively the use of ‘functional-adaptive analyses’ and regarded them only a ‘welcome supplement’ (willkommene Ergänzung, Ax 1988: 84) to the outgroup comparison. Yet, we might think of cases in which such ‘functional-adaptive’ reasons could be helpful. For example, metacentric chromosomes can break and produce two acrocentric chromosomes, or two acrocentric chromosomes can fuse and form one metacentric or submetacentric chromosome. In the case of chromosome fission the centromere must be duplicated prior to fission, and the breakage must happen exactly between the two centromeres in order to reveal two functional acrocentric chromosomes. On the other hand, centric fusion can easily happen. Thus, it is much more likely that in the course of hominoid evolution the chromosome number decreased due to Robertsonian fusion than that it increased by fission (Yunis & Prakash 1982).

Outgroup comparison does not necessarily imply to determine character polarity globally for all characters in the matrix by choosing an outgroup or several outgroups. It is, of course, possible to choose an outgroup for each character individually (as, e.g., in Schmitt 1988). In this manner could a ‘synthetic outgroup’ be composed, consisting of character states in a number of different outgroup taxa (Meier, 1997, suggested a similar procedure; however, he did not explicitly state which taxa were used as outgroups). Naturally, the root can only be attached to this ‘synthetic outgroup’. It would no longer make sense to ‘play around’ with re-rooting networks. Such an approach would circumvent a serious drawback of the ‘outgroup addition’ method: It cannot be expected that the organisms of a concrete taxon chosen as outgroup show all characters in the plesiomorph state. Possibly, the effect of this disadvantage could be minimised by using several outgroups, so that the resulting cladogram would possibly reflect the correct sequence of splitting events. However, if we aim at reaching a comprehensive picture of the phylogeny of a group of organisms, i.e. the evolutionary history consisting of cladogenetic events and anagenetic processes, then possible misinterpretations of character transformation will matter.

This is especially obvious when looking at cladistic analyses based on DNA sequence data exclusively, as, e.g. the one by Nardi et al. (2003). In this study the authors found that Hexapoda are paraphyletic, as Crustacea were closer related to insects than the Collembola. Ironically, they also found that Hymenoptera were sister to Acari, whitch they considered an artefact. Taking also non-molecular characters into consideration, e.g. morphological structures, would have demonstrated

that the presented cladogram is extremely unlikely, notwithstanding the high bootstrap values.

The principle of parsimony pertains to the economy of thinking only. It does not make any statements about nature. Parsimony does not necessarily reflect the economy of nature. We hardly do know anything about the real number of decisions, i.e. possible changes, in the regulation of the ontogenetic development of a character. Thus, just counting ‘steps’ does most probably not tell us anything about the number of non-synonymous mutations leading to a transformation of a character from one state to another. Anyway, there is no rational alternative to the procedure of outgroup comparison. All other ways of polarising characters suggested in the literature are either based on ad-hoc arguments and do, consequently, not provide general tools, or they are based on subjective ideas on plausibility, in the worst case they are nothing but ‘just-so stories’, and consequently are not scientific at all.

Comparing the approaches of Hennig, Ax and the ‘present-day cladists’ shows that there are strong correspondences but also differences (Schmitt in press). In my opinion, the most relevant difference is that the ‘present-day cladists’ no longer discuss character transformations a priori. Determining character polarity globally by rooting is, as far as I see, the crucial disagreement between the ‘traditional Hennigian’ approach and the ‘present-day cladistics’.

Conclusion

Willi Hennig did not establish a method of polarising characters. He decided on the polarity of character states always individually on each character and each taxon. Often he exploited information on the ‘distribution of the characters among the taxa’, which means that he compared putatively closer related taxa, thus applying implicitly an outgroup comparison. Often he argued on the basis of an intuitively assessed plausibility, in some cases even by reference to certain authorities.

Peter Ax accepted the outgroup comparison as the only means to decide on plesiomorph or apomorph condition of a character. He discussed the polarity of characters each by each, and has never published – and probably never compiled – a character matrix.

Modern textbooks of cladistics (Kitching et al. 1998, Schuh & Brower 2009, and Wiley & Lieberman 2011) explain how polarity of characters can be assessed by analysing character by character (as Hennig and Ax proceeded). However, nearly all authors of recent cladistics polarise the characters of their matrices by ‘outgroup addition’ (or ‘outgroup designation’). This

PECKIANA 11 · 2016

41Hennig, Ax, and Present-Day Mainstream Cladistics, on polarising characters

means that rooting between out- and ingroup is the only possibility to convert a network into a cladogram. In analyses based on molecular characters, polarity of characters cannot be assessed in any other way.

As long as we are interested in phylogenesis, and as long as we accept ‘phylogenesis = cladogenesis + anagenesis’, any analysis without considering non-molecular characters is incomplete, if not irrelevant.

Acknowledgements

I cordially thank Stefan Richter (Rostock, Germany) and Gabriele Uhl (Greifswald, Germany) for critically reading the manuscript and giving helpful feedback, Adele Hornig (Oppach, Germany) for permission to use her painting for fig. 1, and Hannes Paulus (Vienna, Austria) for permission to use a photograph of his for fig. 2.

References

Ax, P. (1965): Zur Morphologie und Systematik der Gnathostomulida. – Zeitschrift für zoologische Systematik und Evolutionsforschung 3: 259–276.

Ax, P. (1971): Zur Systematik und Phylogenie der Trigonostominae (Turbellaria, Neorhabdocoela) (Mikrofauna des Meeresbodens 4). – Abhandlungen der Wissenschaften und der Literatur Mainz, mathematisch-naturwissenschaftliche Klasse. Franz Steiner Verlag, Wiesbaden: 84 pp.

Ax, P. (1984): Das Phylogenetische System. – Stuttgart: Gustav Fischer.

Ax, P. (1988): Systematik in der Biologie. – Stuttgart: UTB Gustav Fischer.

Ax, P. (1995): Das System der Metazoa I. Ein Lehrbuch der phylogenetischen Systematik. – Gustav Fischer Verlag, Stuttgart: 226 pp.

Ax, P. (1999): Das System der Metazoa II. Ein Lehrbuch der phylogenetischen Systematik. – Gustav Fischer Verlag, Stuttgart: 383 pp.

Ax, P. (2001): Das System der Metazoa III. Ein Lehrbuch der phylogenetischen Systematik. – Spektrum Akademischer Verlag, Heidelberg: 283 pp.

Beckmann, J. P. (1995): Wilhelm von Ockham. – München: Beck.Carpenter, J. M. (1987): Cladistics of cladists. – Cladistics 3:

363–375.De Jong, R. (1980): Some tools for evolutionary and

phylogenetic studies. – Zeitschrift für zoologische Systematik and Evolutionsforschung 18: 1–23.

Ebach, M. C., J. J. Morrone & D. M. Williams (2008): A new cladistics of cladists. – Biology and Philosophy 23: 153–156.

Hennig, W. (1936a): Uber einige Gesetzmasigkeiten der geographischen Variation in der Reptiliengattung Draco L.: “parallele” und “konvergente” Rassenbildung. – Biologisches Zentralblatt 56: 549–559.

Hennig, W. (1936b): Beziehungen zwischen geographischer Verbreitung und systematischer Gliederung bei einigen Dipterenfamilien: ein Beitrag zum Problem der Gliederung systematischer Kategorien hoherer Ordnung. Zoologischer Anzeiger 116: 161–175.

Hennig, W. (1950): Grundzüge einer Theorie der phylogeneti-schen Systematik. – Berlin: Deutscher Zentralverlag.

Hennig, W. (1966): Phylogenetic Systematics. – Urbana: University of Illinois Press.

Hennig, W. (1968): Kritische Bemerkungen über den Bau der Flügelwurzel bei den Dipteren und die Frage nach der Monophylie der Nematocera. – Stuttgarter Beiträge zur Naturkunde Nr. 193: 1–23.

Hennig, W. (1972): Insektenfossilien aus der unteren Kreide IV. Psychodidae (Phlebotominae), mit einer kritischen Übersicht über das phylogenetische System der Familien und die bisher beschriebenen Fossilien (Diptera). – Stuttgarter Beiträge zur Naturkunde Nr. 241: 1–69.

Hennig, W. (1982): Phylogenetische Systematik. – Hamburg - Berlin: Paul Parey.

Hennig, W. (1983): Stammesgeschichte der Chordaten (Fortschritte der Zoologischen Systematik und Evolutionsforschung 2). – Hamburg - Berlin: Paul Parey.

Hennig, W. (1984): Aufgaben und Probleme stammesgeschicht-licher Forschung. – Berlin-Hamburg: Paul Parey.

Hennig, W. & D. Schlee (1978): Abriß der phylogenetischen Systematik. – Stuttgarter Beiträge zur Naturkunde, Serie A, Nr. 319: 1–11.

Kitching, I. J., P. L. Forey, C. J. Humphries & D. M. Williams (1998): Cladistics. The Theory and Practice of Parsimony Analysis. – 2nd ed., Oxford: Oxford University Press.

Kluge, A. G. & J. S. Farris (1969): Quantitative phyletics and the evolution of anurans. – Systematic Zoology 18: 1–32.

Lundberg, J. G. (1972): Wagner networks and ancestors. – Systematic Zoology 21: 398–413.

Meier, R. (1992): Der Einsatz von Computern in phylogenetischen Analysen - eine Übersicht. – Zoologischer Anzeiger 229: 106–133.

Meier, R. (1995): Advantages and disadvantages of computerized phylogenetic analyses. – Zoologische Beiträge Neue Folge 36: 141–167.

Meier, R. (1997): A test and review of the empirical performance of the ontogenetic criterion. – Systematic Biology 46: 699–721

Nardi, F., G. Spinsanti, J. L. Boore, A. Caparelli, R. Dallai & F. Frati (2003): Hexapod origins: Monophyletic or paraphyletic? – Science 299: 1887–1889.

Nelson, G. J. (1979): Cladistic analysis and synthesis: Principles and definitions, with a historical note on

Michael Schmitt42

PECKIANA 11 · 2016

Adanson’s Familles des Plantes (1763–1764). – Systematic Zoology 28: 1–21.

Nelson, G. J. & N. I. Platnick (1978): The perils of plesiomorphy: Widespread taxa, dispersal, and phenetic biogeography. – Systematic Zoology 27: 474–477.

Nelson, G. J. & N. I. Platnick (1984): Systematics and evolution. – In: Ho, M.-W. & P. T. Saunders (eds): Beyond Neo-Darwinism. – An Introduction to the New Evolutionary Paradigm. – London etc.: Academic Press: 143–158.

Osche, G. (1966): Grundzüge der allgemeinen Phylogenetik. – In: von Bertalanffy, L. & F. Gessner (eds): Handbuch der Biologie Band 3/2. – Frankfurt am Main: Akademische Verlagsgesellschaft Athenaion: 817–906.

Patterson, C. (1980): Cladistics. – Biologist 27: 234–240.Platnick, N. I. (1979): Philosophy and the transformation of

cladistics. – Systematic Zoology 28: 537–546.Remane, A. (1952): Die Grundlagen des natürlichen Systems,

der vergleichenden Anatomie und der Phylogenetik. – Theoretische Morphologie und Systematik I. – Leipzig: Akademische Verlagsgesellschaft Geest & Portig.

Richter, S. & R. Meier (1994): The development of phylogenetic concepts in Hennig’s early theoretical publications (1947–1966). – Systematic Biology 43: 212–221.

Riedl, R. (1975): Die Ordnung des Lebendigen – Systembedingungen der Evolution. – Hamburg - Berlin: Paul Parey (English: Order in living organisms : a systems analysis of evolution. – Chichester - New York: Wiley, 1978).

Schmitt, M. (1988): The Criocerinae: biology, phylogeny and evolution. – In: Jolivet, P., E. Petitpierre & T. H. Hsiao (eds): Biology of Chrysomelidae. – Dordrecht: Kluwer Academic Publishers: 475–495.

Schmitt, M. (2001): Willi Hennig. – In: Jahn, I. & M. Schmitt (eds) (2001). – Darwin & Co. - München: C. H. Beck: 316–343, 541–546.

Schmitt, M. (2013): From Taxonomy to Phylogenetics – Life and Work of Willi Hennig. Leiden – Boston: Brill.

Schmitt, M. (in press.): How much of Hennig is in present-day cladistics? – In: Williams, D. M., Q. D. Wheeler & M. Schmitt (eds): The Future of Phylogenetics – the Legacy of Willy Hennig. – Cambridge, UK: Cambridge University Press.

Schuh, R. T. & A. V. Z. Brower (2009): Biological Systematics – Principles and Applications, 2nd ed. – Ithaca, London: Comstock Publishing Associates.

Sewertsoff, A. N. (1931): Morphologische Gesetzmäßigkeiten der Evolution. – Jena: Gustav Fischer.

Sudhaus, W. & K. Rehfeld (1992): Einführung in die Phylogenetik und Systematik. – Stuttgart: Gustav Fischer.

Throckmorton, L. H. (1968): Concordance and discordance of taxonomic characters in Drosophila classification. – Systematic Zoology 17: 355–387.

Wägele, J.-W. (2000): Grundlagen der Phylogenetischen Systematik. – München: Dr. Friedrich Pfeil (English: Foundations of phylogenetic systematics. – München: Pfeil, 2005).

Watrous, L. E. & Q. D. Wheeler (1981): The out-group comparison method of character analysis. Systematic Zoology 30: 1–11.

Wiley, E. O. (1981): Phylogenetics. – New York etc.: John Wiley and Sons.

Wiley, E. O. & B. S. Lieberman (2011): Phylogenetics - Theory and Practice of Phylogenetic Systematics, 2nd ed. – Hoboken, NJ: Wiley-Blackwell.

Willmann, R. (1989): Paleontology and the systematization of natural taxa. – Abhandlungen des Naturwissenschaftlichen Vereins in Hamburg Neue Folge 28: 267–291.

Yunis, J. J. & O. Prakash (1982): The origin of man: A chromosomal pictorial legacy. – Science 215: 1525–1530.